You are only browsing one thread in the discussion! All comments are available on the post page.

Return

Ertebolle ,

Yep. It sucks that this is the choice we have to make, but it is, at least until we figure out a way to fix our voting system.

chaogomu ,

STAR voting would fix everything. This website goes into more detail.

The simplified science of it all says that if a voting method forces you to choose between candidates or rank them in a fixed order, then that very ranking will, over time, promote two dominant parties. (Arrow's Theorem)

A cardinal voting system, such as STAR, is immune to Arrow's Theorem. STAR was designed to be the absolute best voting system possible. It's easy to use, easy to count, and gives better results than any other system.

Ertebolle ,

I'm a fan of MMP voting (used in Germany, New Zealand, and the Scottish/Welsh parliaments in the UK), where you vote for a district representative and a party and the parties get extra seats to ensure that the proportions balance out. It's easy for voters to understand - no ranking or rating or whatever - and it simultaneously lets you support third parties (because they'll get some seats even if they don't have a majority in any district), lets you vote for the best candidate in your particular district without regard to their party (if you like your local Republican but you hate national Republicans you can simultaneously vote for your guy + for him to be in the minority), eliminates gerrymandering (since party representation comes from a percentage of the overall vote), and makes every vote count (since even in a deep red district your blue vote still contributes to the national total for your party and therefore its share of legislative seats).

If that proves successful then we can explore other systems for national/presidential votes, but you're never going to get a serious third-party movement in the US if you insist on starting with the White House - reforms to support third-party presidential candidates are the sort of thing you do after you've got 40 or 50 minor party representatives in Congress.

chaogomu ,

The problem with multi-member districts (which are required for proportional voting) is the fact that to get rid of an incumbent, you need a vast majority to actively vote against them. For example, in a 5 member district, you need over 80% of the vote against one bad incumbent to get rid of them.

Proportional voting also explicitly makes political parties part of government. The goal is to not do that.

Ertebolle ,

MMP doesn't require multi-member districts; the extra seats are not tied to a particular district, they draw from a list of names submitted by each party.

Zagorath ,
@Zagorath@aussie.zone avatar

Except for the fact that any cardinal system other that Approval is absolutely trivial to game, automatically devolving into Approval.

chaogomu ,

STAR gets around that by adding the runoff step.

Also, Approval gives better results than any Ordinal system, because Approval is also immune to Arrow's Theorem.

Zagorath ,
@Zagorath@aussie.zone avatar

Every voting system has pros and cons. It’s impossible to create a perfect voting system because there are multiple mutually-exclusive criteria by which a system can be measured.

So it’s important in such discussions to be forthright about which criteria you consider more or less important than others.

In order to avoid the spoiler effect and to discourage trivial tactical voting. To these ends, criteria like the Later-No-Harm criterion and Favourite Betrayal are important, but one need also look at the ways in which systems that fail them do so.

Approval Voting fails LNH trivially. In a genuine three-candidate race (i.e., one where prior to the election, you know all three have a genuine chance of winning) where your honest vote is to approve of Left and Centre and to disapprove of Right, but where you more strongly approve of Left, you are disincentivised from voting honestly because that will hurt the chances of Left winning. Unfortunately in so doing you also increase the chances of Right winning, compared to if you voted honestly. Basically, you’re forced to make a decision between your honest vote which increases the chances of a mediocre result, and a dishonest vote which increases the chances of either a very good or very bad result. It’s LNH because you’re incentivised to not vote for Centre even though you honestly would have.

In IRV you get a similar outcome in theory, but what we’ve seen in practice is that it doesn’t actually play out. The difference comes down to how preferences are distributed and who gets eliminated. If Centre ends up coming last on first preferences, that’s where Favourite Betrayal comes in. In your honest vote, Centre’s preferences distribution is entirely up to those who voted Centre, which could be a mix of Left and Right, which risks Right winning. If you had voted dishonestly and put Centre ahead of Left, you’re increasing the chance that Centre isn’t eliminated first, and instead Left is, with Left votes going to Centre—a better outcome than Right winning.

But the thing is, in practice this doesn’t tend to be the case. I live in a seat where this happened in our election last year, and it turns out that people who vote for one candidate overwhelmingly tend to second preference the same candidate (at least once smaller non-viable candidates are ignored). The Australian Greens (Left) and the Australian Labor Party (Centre) actually preference each other at the same rate. And so it came down to the fact that Labor was eliminated first (after non-viable candidates) by a very narrow margin, giving almost all their next preferences to the Greens, resulting in a Greens win. Not voting strategically, in the real world, actually pays off under IRV. Under Approval, because the ability to actually express your nuanced preference doesn’t exist, tactical voting is more strongly encouraged. In summary, while LNH and Favourite Betrayal are, in a sense, “equally bad”, the former is more of a problem when looking at the real-world preferences of voters, and a system which fails the latter should be preferred over one that fails the former.

chaogomu ,

IRV is a terrible system.

Later-no-harm is sort of a meaningless criterion that was invented by a group called Fairvote to push IRV under the name Ranked Choice.

They invented it to say that Cardinal voting systems don't let you rank preferences. Which is sort of the entire point of cardinal systems. It's not much of an issue in the real world, because if you're happy with A or B, then you put down a vote for A and B. In any large scale election, there will be enough people who have a set preference that they only chose A or B and not both. The point being, a vote for one does nothing to impact a vote for the other, because you count the votes independently of each other.

The problems of IRV are many and varied. It has to be counted in a single centralized location, which leads to problems and security issues. It still has favorite betrayal, and the more viable candidates you have, the worse it gets. This means that you have to have some strategy while voting, but it's much harder for the average voter to know if their strategy will do any good.

Then there's the issue of exhausted ballots. Again, the more viable candidates you have, the worse it gets. Most of my data is from US elections that use the system, but the city of San Francisco sees about 18% of ballots thrown out due to ballot exhaustion. That alone is horrific.

Now, switching to another plot here, STAR is not approval. Star lets you voice your true preferences for candidates in a way that Approval and even IRV do not. Take a look at this graphic again.

First off, the average person sees it and says oh, it's a 5-star review. I know how those work. Then they rate each candidate on the scale of 0-5 stars. That's actually one of the most common ways people fuck up an IRV ballot. They think it's a 5-star review and give multiple candidates the same number.

So the next part of STAR is the automatic runoff. You take the two highest scored candidates and then put them head to head, but you use the preferences on the ballots to do it. If A is rated higher than B on that ballot, then the vote goes to A. If A and B are the same rating, then that ballot is counted as "No Preference". And the number of those ballots is also released at the end.

STAR gives you so much more information about candidates than any other system. You have an instant approval rating in the form of a 5-star average for every candidate.

Zagorath ,
@Zagorath@aussie.zone avatar

IRV is a terrible system.

Later-no-harm is sort of a meaningless criterion that was invented by a group called Fairvote to push IRV

Oh sorry, I thought we were here having a civil conversation about an interesting and complicated subject. If you’re going to start off with that sort of tribalistic bad faith bullshit I’m out.

chaogomu ,

IRV is a terrible system, I know in another comment you said it was your favorite, but it's almost as bad as First Past the Post, and actually worse in a few areas.

And seriously, Later-no-harm is meaningless. It's basically a "did you vote for this person? They might win". It's like, no shit, that's how elections work.

If you want an actual issue with elections, look at the Monotonicity criterion. IRV fails this one. You can actually almost guarantee your most hated candidate wins, by voting for your favorite in the first round.

This brings up the issue of ballot exhaustion again. If your first round pick survives multiple rounds before being eliminated, your ballot suddenly doesn't have any valid candidates on it. This means your ballot might as well have been empty.

If you had not voted for your favorite, your vote could have gone to one of the others and helped them win, instead no, all that information about your preferences is just thrown out because of a stupid, arbitrary rule. If every single voter puts Candidate B as their second choice, Candidate B has 100% approval, and yet, under IRV, Candidate B is the first eliminated.

Another issue. Ranking candidates in order tells us nothing about how you actually feel about them. We know you like number 1, but number 2 could be anything between Jesus and Hitler. There's no information there, just that you like them less than 1.

Zagorath ,
@Zagorath@aussie.zone avatar

it was your favorite

It’s my favourite single-winner, because I’ve looked at all the others and decided the downsides of the others outweigh the downsides of IRV. But I’d still rank any proportional system much higher than IRV. And ironically, STV (which is a multi-winner version of IRV) is among my least favourite, due to its relatively limited almost quasi-proportional nature.

Later-no-harm is meaningless. It’s basically a “did you vote for this person? They might win”

Well, no. It’s that by adding more candidates to your ballot who were not your favourite choice, you could actually decrease the chance your favourite candidate wins. It’s that a tactical dishonest vote can be more optimal than an honest vote. And that’s bad. That’s bad in the real world.

I know IRV fails in a lot of theoretical ways. But when it’s used in the real world that just doesn’t matter. I mean, it could theoretically matter, but with how real people actually vote, it doesn’t. I explained how that happens earlier.

This brings up the issue of ballot exhaustion again

Definitely a problem. And unfortunately I’ve seen it matter, with some candidates dishonestly promoting “just vote 1” in the closest thing Australia has to the type of voter discouragement campaigns that are so rife in America. Federally and in my state, Australia uses compulsory preferential voting. You have to number every candidate. This eliminates the exhausted ballot problem. Our local council elections are where exhaustion becomes a problem. The solution isn’t to move to an inferior voting system, it’s to use the same compulsory preferential system used in other elections.

We know you like number 1, but number 2 could be anything between Jesus and Hitler. There’s no information there, just that you like them less than 1.

This is a feature, not a bug. The fact that you’re thinking about it as a disadvantage says a lot to me about why you like cardinal systems. I fundamentally disagree.

It doesn’t matter if I love one candidate, like another, and hate the third, or like one, dislike another, and hate the third. What matters is who my vote helps elect. And I want the first one to win, or if they can’t, I want the second one to win. And that’s what IRV perfectly represents. In a cardinal system, if I vote 5, 2, 1, as is my honest preference, all that does is help my least favourite candidate win if my favourite doesn’t, compared to if I voted 5,5,1, or 5,4,1. That’s how any cardinal system inevitably devolves into approval. And again with approval, I lose the ability to distinguish preference. You say it’s bad that my second preference could be Jesus or Hitler, but at least with IRV I can clearly say I like Jesus more than Hitler, instead of just saying I “approve” of both because the only remaining option is Pol Pot. At the point where I know Jesus isn’t going to win, it doesn’t matter how much more I prefer Hitler over Pol Pot. I just want to ensure Pol Pot doesn’t win. Ordinal voting better represents how a rational voter thinks about the candidates than cardinal voting does. And that’s why it’s better.

chaogomu ,

Okay, you take two of the worst parts of IRV and pretend they're somehow good. That's mind-boggling.

Ballot exhaustion is not solved by compulsory preferential voting. It only hides the fact that you now have to rank all the candidates. So when your middle preferences are eliminated before your first is, you've now been forced to elect your most hated option.

And again, later-no-harm is still a “did you vote for this person? They might win” criterion. Because if you don't like someone, don't vote for them.

Finally, you still have no clue how STAR works. It's not 1-5. It's 0-5. And the Automatic runoff part is pretty important, that part of it means that your vote goes to the finalist who you rated higher, not to merely the person who got the most points in the first round.

Zagorath ,
@Zagorath@aussie.zone avatar

I think one other important detail though. While IRV is my preferred single-winner system, I fundamentally think that single-winner systems are flawed. Elections should avoid them as much as possible, in favour of proportional multi-winner systems.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • random
  • [email protected]
  • All magazines
    •